def __init__(self, vocab_size, d_model, N, heads, dropout, 最大长度=1024):
super().__init__()
self.N = N
self.embed = Embedder(vocab_size, d_model)
self.embedP = Embedder(最大长度, d_model)
# self.pe = PositionalEncoder(d_model, dropout=dropout)
self.layers = get_clones(DecoderLayer(d_model, heads, dropout), N)
self.norm = Norm(d_model)
def __init__(self, vocab_size, d_model, N, heads, dropout):
super().__init__()
self.N = N
self.embed = Embedder(vocab_size, d_model)
#self.pe = PositionalEncoder(d_model=47, max_seq_len=13,dropout=dropout)
self.layers = get_clones(EncoderLayer(d_model, heads, dropout), N)
self.norm = Norm(d_model)
def __init__(self, vocab_size, d_model, N_layers, heads, dropout):
super().__init__()
self.N_layers = N_layers
self.embed = Embedder(vocab_size, d_model)
self.pe = PositionalEncoder(d_model, dropout=dropout)
self.layers = get_clones(DecoderLayer(d_model, heads, dropout),
N_layers)
self.norm = Norm(d_model)
def __init__(self, vocab_size, d_model, N, heads, dropout, c, n_layers):
super().__init__()
self.N = N
self.embed = Embedder(vocab_size, d_model)
self.pe = PositionalEncoder(d_model, dropout=dropout)
self.dynam = Ucb_bandit(n_layers, c)
self.layer = EncoderLayer(d_model, heads, dropout)
self.layers = get_clones(EncoderLayer(d_model, heads, dropout), N)
self.norm = Norm(d_model)
def __init__(self, vocab_size, d_model, N, heads, dropout, field, word_emb,
opt):
super().__init__()
self.N = N
self.word_emb = word_emb
self.opt = opt # unused, just for querying
self.embed = Embedder(vocab_size, d_model, word_emb, field)
self.pe = PositionalEncoder(d_model, dropout=dropout)
self.layers = get_clones(EncoderLayer(d_model, heads, dropout),
N) # attention
self.norm = Norm(d_model)
